[00:52:23] (03CR) 10Legoktm: [C: 032] Remove backward usage of ApiBase::dieUsageMsg [extensions/ORES] - 10https://gerrit.wikimedia.org/r/453587 (owner: 10Umherirrender) [00:58:29] (03Merged) 10jenkins-bot: Remove backward usage of ApiBase::dieUsageMsg [extensions/ORES] - 10https://gerrit.wikimedia.org/r/453587 (owner: 10Umherirrender) [01:14:17] (03CR) 10jenkins-bot: Remove backward usage of ApiBase::dieUsageMsg [extensions/ORES] - 10https://gerrit.wikimedia.org/r/453587 (owner: 10Umherirrender) [04:58:55] 10Scoring-platform-team, 10DBA, 10MediaWiki-Database, 10Blocked-on-schema-change, 10User-Ladsgroup: Schema change for rc_this_oldid index - https://phabricator.wikimedia.org/T202167 (10Marostegui) MySQL doesn't support index renaming, we'd need to drop+create anyways. [04:59:13] 10Scoring-platform-team, 10DBA, 10MediaWiki-Database, 10Blocked-on-schema-change, 10User-Ladsgroup: Schema change for rc_this_oldid index - https://phabricator.wikimedia.org/T202167 (10Marostegui) p:05Triage>03Normal [09:17:25] o/ [10:28:19] 10Scoring-platform-team, 10Analytics, 10ORES, 10Services (designing): ORES hooks - https://phabricator.wikimedia.org/T201869 (10Ladsgroup) There is one thing: ORES extension hooks into RC changes and not revision creations so it can't have such hook but it can be something like this: ``` /** * Called aft... [10:59:13] halfak: around? [10:59:29] Yes. Partially. I'm packing up :) [11:00:03] https://github.com/organizations/wiki-ai/settings/billing [11:00:14] can you see https://github.com/organizations/wikimedia/settings/billing ? [11:01:00] Yes [11:01:10] Oh wait. Let me check the WMF one [11:01:22] No I cannnot [11:01:24] wikimedia org. has some coupons that allows it to have more repos and data [11:02:31] It's almost impossible to see what is their git lfs limit if there any [11:03:30] what is obvious is that they are in better shape than wiki-ai given they have coupons [11:04:35] halfak: what do you think? [11:06:33] It means probably we can enable git lfs in wikimedia org [11:06:38] Hmm.. Doesn't seem like there is any LFS [11:06:55] But yeah, definitely in better shape than wiki-ai. [11:07:05] We need to figure out what it would take to get LFS. [11:07:26] It might mean finding budget. It might mean getting github to donate us some LFS. [11:07:57] I need to check [11:08:04] Amir1, you could run a test. Try to create one of our big repos in wikimedia/ [11:08:09] And convert it to LFS. [11:08:25] There's a nice conversion utility that I documented in my last attempt [11:08:44] You need to convert it on your laptop and then push the brand new repo to a new remote. [11:09:06] I've got to run to get back to packing. I should be back in a few hours. [11:09:20] halfak: before you go, can you link me? [11:09:24] the utility [11:19:04] I'll need to dig through phab. It's an old task in our backlog with "LFS" in the title. [11:19:07] Amir1, ^ [11:20:54] https://phabricator.wikimedia.org/T181678#3798104 [11:21:00] Looks like I was working from awight's notes! [11:22:26] Thanks! [13:13:51] halfak: tell me when you're back [13:19:00] akosiaris: Welcome back [13:19:46] 10Scoring-platform-team (Current), 10User-Ladsgroup: [Epic] Use LFS for large ORES files - https://phabricator.wikimedia.org/T197096 (10Ladsgroup) a:03Ladsgroup Already on it. [13:20:15] 10Scoring-platform-team (Current), 10User-Ladsgroup: Migrate models to LFS - https://phabricator.wikimedia.org/T197097 (10Ladsgroup) a:03Ladsgroup [13:20:28] Amir1: o/ [13:20:32] 10Scoring-platform-team (Current), 10User-Ladsgroup: Confirm that continuous integration is compatible with LFS - https://phabricator.wikimedia.org/T197605 (10Ladsgroup) a:03Ladsgroup [13:21:16] akosiaris: I have lots of things to discuss with you, let me know when you are a little bit free, I guess you're already drown in emails and phab tickets [13:23:27] Amir1: exactly. I 'll ping you sometime this week [13:25:21] 10Scoring-platform-team (Current), 10Release-Engineering-Team (Watching / External), 10User-Ladsgroup: Another round of discussion about wiki-ai's GitHub->gerrit mirroring - https://phabricator.wikimedia.org/T194212 (10Ladsgroup) a:03Ladsgroup I've enabled and was able to use git-lfs in wikimedia org in gi... [13:31:42] 10Scoring-platform-team, 10ORES: Consider committing all non-private datasets to the repos - https://phabricator.wikimedia.org/T192617 (10Ladsgroup) No, we should not do this. Git LFS actually doesn't allow us to store large text files, it performs very poorly in these cases because in git-lfs objects don't ge... [13:52:03] https://github.com/wikimedia/articlequality [13:52:10] It works, it fucking works \o/ [13:52:54] wikimedia/articlequality#1 (master - 73e7aa9 : Aaron Halfaker): The build passed. https://travis-ci.com/wikimedia/articlequality/builds/82307981 [13:53:17] haha The CI also works ^ [13:53:28] * Amir1 is over the moon [13:58:43] o/ Amir1 [13:58:50] Back just in time for our 1:1. [13:59:02] yup [13:59:06] Getting my damn bike packed was ridiculous! [14:04:16] 10Scoring-platform-team, 10DBA, 10MediaWiki-Database, 10Blocked-on-schema-change, 10User-Ladsgroup: Schema change for rc_this_oldid index - https://phabricator.wikimedia.org/T202167 (10Anomie) I think they added index renaming in MySQL 5.7. I don't know if MariaDB picked it up yet ([[https://jira.mariadb... [15:13:37] 10Scoring-platform-team, 10DBA, 10MediaWiki-Database, 10Blocked-on-schema-change, 10User-Ladsgroup: Schema change for rc_this_oldid index - https://phabricator.wikimedia.org/T202167 (10jcrespo) I am not saying this is needed, but technically the safe way to rename an index (without T202167#4514755) is to... [15:29:14] I need to grab something before the meeting [15:29:20] otherwise I will starve :D [15:29:27] will be back soon [15:46:22] Change of plans, I couldn't make to the the sync meeting if I try to eat anything [15:46:31] I will eat after the sync meeting [15:48:02] https://etherpad.wikimedia.org/p/scoring_current_work [15:49:37] heads up, I'm moving all repos to wikimedia org in github [15:49:54] Aaron-approved (TM) [15:54:12] ORES is moved [16:06:52] so all ORES is on GitHub or in Gerrit now? [16:07:08] so I don't mess up with the odd patches I might submit in the future [16:07:19] Github [16:07:23] but another org [16:07:48] oki :) [16:24:58] 10Scoring-platform-team (Current), 10Wikilabels: Extend wikilabels to support session-labelling - https://phabricator.wikimedia.org/T201370 (10notconfusing) The PR is merged. Next to do is to load make test data. Questions to answer for test data that I'm pondering: 1. Which wiki? 2. I'm thinking newcomers. Wh... [16:55:13] afk for lunch [17:05:12] Me too! [17:20:54] back now, I have a hate-love relationship with Doner but it works [17:37:27] 10Scoring-platform-team (Current), 10Release-Engineering-Team (Watching / External), 10User-Ladsgroup: Another round of discussion about wiki-ai's GitHub->gerrit mirroring - https://phabricator.wikimedia.org/T194212 (10Ladsgroup) Everything is moved now: https://github.com/orgs/wikimedia/teams/scoring-platfo... [17:38:16] 10Scoring-platform-team (Current), 10User-Ladsgroup: Migrate models to LFS - https://phabricator.wikimedia.org/T197097 (10Ladsgroup) I'm reshaping all of the repos to use git lfs. Took a back up https://analytics.wikimedia.org/datasets/archive/public-datasets/all/ores/20180820/ and also will keep another backu... [18:00:35] back too [18:02:50] halfak: first, https://github.com/wiki-ai/ These repos are left [18:03:23] some need to be archived I think, some need to be moved but I don't feel like moving the experimental repos [18:03:38] These are the moved ones: https://github.com/orgs/wikimedia/teams/scoring-platform/repositories [18:03:56] Also I want to archive https://github.com/wiki-ai/itemquality [18:04:36] archived [18:05:03] halfak: awight please review https://github.com/wikimedia/ores/pull/258/files before I move on to wiring the whole thing [18:05:27] and the articlequality now is fully LFS, moving on to the next one [18:07:35] That’s exciting! Now we need strategize the GitHub storage costs, and how to migrate our models [18:07:41] & whether we want to include datasets [18:08:10] I'm already migrating the models, spent most of today trying to figure this out :D [18:08:22] I don't think we should have all of the datasets [18:08:54] git lfs doesn't compress anything and size of the repo will sky rocket if we put non-binary things there [18:09:16] git itself is pretty good at compressing things it can delta [18:09:34] Our datasets should be gzipped… [18:09:48] draftquality is fully lfs now, migrating to drafttopic [18:10:08] awight: That's valid but still lots of work IMO, I can be convinced otherwise [18:10:10] I agree that they probably shouldn’t be in the main repos, but it would make sense if we could track and download optionally. [18:10:40] The motivation is that extraction takes a *long* time, of course [18:12:07] drafttopic is done now, moving to editqaulity [18:13:03] Either way, we can look at datasets later, it shouldn’t block the work you’re doing now [18:13:08] Thank you for taking this on! [18:13:21] <3 [18:17:49] Hi, Geeks [18:19:37] It's uploading editquality which is 5.1GB and that's going to take some time [18:22:07] awight: Also I have this for you :D https://gerrit.wikimedia.org/r/c/mediawiki/extensions/ORES/+/452990 [18:23:52] will do [18:25:36] (03CR) 10Awight: [C: 032] "Good cleanup!" [extensions/ORES] - 10https://gerrit.wikimedia.org/r/452990 (https://phabricator.wikimedia.org/T201918) (owner: 10Ladsgroup) [18:29:35] And this straightforward merge: https://github.com/wikimedia/ores/pull/259 [18:32:12] (03Merged) 10jenkins-bot: Do not try to cache scores that have been requested via API [extensions/ORES] - 10https://gerrit.wikimedia.org/r/452990 (https://phabricator.wikimedia.org/T201918) (owner: 10Ladsgroup) [18:34:48] Aand https://github.com/wikimedia/wikilabels/pull/244 [18:36:12] While we are here: https://github.com/wikimedia/ores-support-checklist/pull/6 :D [18:36:17] I'm horrible, I know [18:36:27] (03CR) 10jenkins-bot: Do not try to cache scores that have been requested via API [extensions/ORES] - 10https://gerrit.wikimedia.org/r/452990 (https://phabricator.wikimedia.org/T201918) (owner: 10Ladsgroup) [18:37:17] how dare you :p [18:47:50] o/ [18:48:00] Just came up for air from meetings. [18:48:17] sounds terrible [18:48:57] Yes. But good stuff still. Just talked to Corey and Ian about the curation vs. DBAs. [18:49:17] It seems there a good strategic plan point brewing WRT figuring out what to do. [18:49:27] awight: one thing: Is wheels are migrated to lfs? [18:49:28] wikimedia/wikilabels#402 (fix_links - bd1e63d : Amir Sarabadani): The build failed. https://travis-ci.org/wikimedia/wikilabels/builds/418361519 [18:49:39] Amir1: I wouldn’t bother [18:49:58] Corey is thinking about building up a "Storage Team" to try to tackle this stuff. I think that's great news because if he's successful we can get out of the game of popularizing our difficult set of tradeoffs with JADE. [18:53:45] 10Scoring-platform-team (Current), 10User-Ladsgroup: Confirm that continuous integration is compatible with LFS - https://phabricator.wikimedia.org/T197605 (10Ladsgroup) It is: https://travis-ci.org/wikimedia/ores [18:53:53] Sounds like the biggest controversy will be in selecting the team [18:57:23] 10Scoring-platform-team (Current), 10User-Ladsgroup: Confirm that continuous integration is compatible with LFS - https://phabricator.wikimedia.org/T197605 (10awight) @Ladsgroup I'm not sure this makes the case—enabling LFS isn't going to trigger any incompatiblities, the trick is that we need to have the LFS... [19:01:08] halfak: do you have any good links/backstory for the Basque Wikipedia partnership? [19:01:35] oh, I'm also very curious about this storage team proposal. [19:05:46] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10Ladsgroup) 05Resolved>03Open After migrating the github repos to use git lfs, now it returns this error: ``` amsa@C235:~/ores-prod-dep... [19:07:34] 10Scoring-platform-team (Current), 10Release-Engineering-Team (Watching / External), 10User-Ladsgroup: Another round of discussion about wiki-ai's GitHub->gerrit mirroring - https://phabricator.wikimedia.org/T194212 (10Ladsgroup) All of the repos with model files now are using git lfs [19:08:02] 10Scoring-platform-team (Current), 10User-Ladsgroup: Migrate models to LFS - https://phabricator.wikimedia.org/T197097 (10Ladsgroup) All of the repos with model files now are using git lfs. Deploying this to prod is blocked on T192042 [19:15:31] I go home, will a little bit more from home [19:35:49] 10Scoring-platform-team, 10Analytics, 10ORES, 10Services (designing): ORES hooks - https://phabricator.wikimedia.org/T201869 (10Pchelolo) > ORESRecentChangeScoreSaved( RecentChange $rc, array $scores ) This hook will be executed from the `FetchScoreJob` since that's where we have the scores to pass into i... [19:51:28] Amir1: Any ORES patches pending deployment? [19:52:58] Amir1: I see you’re commuting, but if you see this backscroll I wanted to set up one more chat before Wednesday, to continue syncing about the alternative implementations and risks. [19:56:13] awight: do you think we're all set for Wednesday's RFC meeting? [19:58:38] harej: Mostly, though I’d still like to talk through any remaining concerns [19:59:10] The “log” implementation is dramatically weird IMO, so I want to understand the motivation better. [20:00:15] It's just an alternative that doesn't use the revision table. I kind of liked the idea of judgments as events that happen and are logged, but seeing the spreadsheet spell out the implementation challenges made me less enthusiastic about it. [20:00:57] I think it falls apart when you examine how an edit would work [20:01:31] However, it begs an important question of what the editing will look like [20:02:27] is this people going in to fix typos or elaborate on their own judgments? Where do replies to judgments happen? Is there a use case for bulk editing many judgments? [20:03:25] Seems like we need to get the most *general* implementation deployed in the wild, so we can see what actually happens. [20:04:54] And pages are the least weird way to do that [20:11:09] hehe, that’s a good way to put it [20:11:35] It’s something we can do right away, that gives us the features we need for a public release [20:25:19] awight: I think there is one, the poolcounter [20:25:23] awight: Sure thing [20:26:41] Just make sure it doesn't conflict with shit ton of meetings I have [20:29:40] Amir1: ah cool, I’ve missed the window for today but could deploy that to beta [20:29:42] what does the poolcounter have to do with ores? [20:30:09] awight: don't worry. I don't think we should deploy anything to prod now [20:30:36] first I need to make sure the git lfs thingy works, right now it's a little bit weird (It doesn't get cloned to gerrit...) [20:30:41] Platonides: This would be to prevent massive, parallel ORES requests from a single IP. [20:31:06] awight: and for the poolcounter, this one is the class, I haven't wired it to the rest of the ores :D [20:31:58] ah hehe mebbe we won’t deploy yet [20:32:40] awight: do you think you can take a look at this? https://github.com/wikimedia/ores/pull/258 [20:33:33] oh, you are using a python client to poolcounter [20:33:44] Amir1: looks reasonable, I’ll let you address tgr’s comment [20:35:24] tgr is right [20:35:41] if you get disconnected, you lose the lock [20:37:01] It's true but it might happen that we try to release on non-existent connection than everything falls apart and it starts to error out. It's more of a sanity check IMO [20:39:19] well, if you connect just to send a release [20:39:27] do note that the daemon won't return RELEASED [20:39:32] but NOT_LOCKED [20:41:19] Yup and then " return data == 'RELEASED\n'" will return false which means it's not released and then the ores can handle it gracefully [20:41:19] hmm [20:41:26] without exploding :D [20:41:35] if not self.stream: [20:41:39] return False [20:41:44] would do it, too ;) [20:41:52] hmm, yeah [20:42:20] I'm wondering [20:42:23] where will you call it? [20:42:51] perhaps using ACQ4ANY would make more sense than ACQ4ME in some points? [20:44:19] Platonides: I applied your idea [20:44:41] I looked at which one, I think ACQ4ME makes sense in our case [20:45:45] ok [20:46:04] 10Scoring-platform-team (Current), 10User-Ladsgroup: Confirm that continuous integration is compatible with LFS - https://phabricator.wikimedia.org/T197605 (10Ladsgroup) Oh I see, the thing is we don't run any tests on Zuul CI right now (except on our mediawiki extensions but that's not the case here) so I don... [20:53:52] Amir1: sorry for reviewing in “single comment” mode, hopefully you aren’t getting a spamflood [20:54:19] nope, don't worry I was fixing them [20:55:30] awight: Do you have any interface documentation handy in ores? [20:58:52] Or literally, for an interface? [20:59:12] literally for an interface [20:59:24] I added some doc already, I hope that's fine for now :P [21:00:55] okay awight can you tell what the remaining issues are now? [21:01:06] the name of the method, the config part? [21:01:59] awight: the thing about the function variables are that it might be need different values for different usecases [21:02:38] for example I can argue we can use this type of locking for when a worker wants to score a revision and it doesn't want other workers to do it [21:03:06] that's exactly what poolcounter is made for but that type of locking requires different number of workers and timeout, etc. [21:03:15] That’s fine, the calling function can read from config in that case ;-) [21:03:26] We already have a mechanism for the case you just described, btw [21:03:36] yup, that's my point [21:04:12] regarding that, it was just an example. I know we have that. PoolCounter might prove to be so efficient we switchover to that too :D you never know :D [21:04:30] The celery queue keys only vary with request params for job deduplicating (said mostly for the benefit of any innocent bystanders.) [21:04:59] hell yeah, well PoolCounter is an orthogonal mechanism for now, and it makes sense. [21:05:45] The user’s IP is unrelated to whether we’ve seen this scoring request before. [21:09:36] awight: one thing, I've been trying to update submodule in the ores deploy and I get this error: https://phabricator.wikimedia.org/T192042#4515985 Is it anything obvious that I'm missing? [21:09:57] “smudge”… error… ? [21:10:30] I’ll check for that reference in my repo [21:12:57] awight: regarding print(), That was debug leftover :D I don't want to put debug in it for now [21:14:59] My ores-deploy#master is 171212271e77db220c022d416c6b7adefefc4d9d, submodules look fine [21:15:18] re-clone? [21:15:24] try unplugging and plugging back in ;-) [21:16:14] I think I’m not recursing into submodules with lfs [21:18:32] Gross, “git lfs help fetch” doesn’t pipe through my $PAGER [21:20:34] git submodule foreach --recursive 'git lfs pull’, if that’s a thing we have to do now? [21:21:34] Amir1: *ah* I’ve been in ores-deploy#master this whole time, no wonder I wasn’t seeing the LFS work. [21:21:53] checking articlequality#master now. [21:25:10] I'm calling it a day, see you on Wednesday o/ [21:25:38] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10awight) @Ladsgroup Please be sure to share the update procedure with us once you're ready. I get this warning: > Your branch and 'origin/... [21:26:02] Have a good night, what’s left of it! [21:29:11] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10Ladsgroup) Yeah, The thing is the whole history is rewritten now, so the hashes of commits are different and now your system thinks they a... [21:29:38] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10awight) I get the same error: ``` $ git status On branch master Your branch and 'origin/master' have diverged, and have 251 and 251 diffe... [21:32:56] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10awight) Here's the error log mentioned above, {F25187675} [21:36:49] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10Ladsgroup) >>! In T192042#4516643, @awight wrote: > In the future it seems prudent to do this sort of rewrite into a named branch, but I g... [21:37:17] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10Ladsgroup) FYI, completely removing and adding a submodule didn't work either. [21:37:47] (03CR) 10jenkins-bot: Localisation updates from https://translatewiki.net. [extensions/ORES] - 10https://gerrit.wikimedia.org/r/454125 (owner: 10L10n-bot) [21:39:40] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10awight) >>! In T192042#4516665, @Ladsgroup wrote: >>>! In T192042#4516643, @awight wrote: >> In the future it seems prudent to do this sor... [21:53:17] 10Scoring-platform-team (Current), 10User-Ladsgroup: Confirm that continuous integration is compatible with LFS - https://phabricator.wikimedia.org/T197605 (10awight) Good point about Zuul, it does seem harmless now that you point that out. Also, it's nice that you demonstrated we can force push to the master... [21:57:27] 10Scoring-platform-team, 10ORES, 10Services (designing): Merge ORES precaching with ORESFetchScoreJob - https://phabricator.wikimedia.org/T201868 (10Ladsgroup) >>! In T201868#4511507, @Pchelolo wrote: > Sorry for bothering again, I might have misunderstood - how does this compare to what `/precache` endpoint... [21:59:06] 10Scoring-platform-team, 10Analytics, 10ORES, 10Services (designing): ORES hooks - https://phabricator.wikimedia.org/T201869 (10Ladsgroup) >>! In T201869#4516098, @Pchelolo wrote: > This hook will be executed from the `FetchScoreJob` since that's where we have the scores to pass into it, and the RC id is n... [22:02:26] 10Scoring-platform-team, 10ORES, 10Services (designing): Merge ORES precaching with ORESFetchScoreJob - https://phabricator.wikimedia.org/T201868 (10Pchelolo) > It is way more complex, precache end point is on the server side, this is client and only hit the ores service when needed for example it only sends... [22:15:08] harej: I’m going to state some of our use cases, https://docs.google.com/spreadsheets/d/1RPb8VHbseE_xPe46nFqo4QVYmwzgfFJrO4_Wh-QKBSw/edit#gid=0 [22:19:40] awight: can you get me access to that document? [22:19:51] oops [22:21:33] {{done}} [22:21:33] How efficient, awight! [22:22:18] awight: I'm pushing the new history to the gerrit. It should be done soonish, my home connection is not great for upload so it takes some time [22:22:41] 2% done, lovely <3 [22:22:57] I'm leaving this time for real [22:22:58] getting your money’s worth from the ISP, then! [22:23:05] o/ [22:23:09] :P [22:23:46] I’ve been using a hippie ISP for years, they can rarely get me 1Mbps, it’s downright embarrassing. [22:24:27] Luckily, at that low of a speed nobody seems to care about DMCA issues ;-) [22:24:44] Truly at those speeds your patience is rewarded. [22:25:08] yah no impulse-buying happening in this household. [22:26:03] only thing is, if the Hollywood secret police do eventually catch up with me, I can’t argue that it wasn’t premeditated [22:26:59] ty for the edits! [22:27:51] 10Scoring-platform-team, 10JADE: Bikeshed: Explore alternative names for "judgment" in JADE - https://phabricator.wikimedia.org/T200365 (10kaldari) Please don't call it "Review"! There are already too many things on Wikipedia called review: the page reviewing process associated with https://en.wikipedia.org/wi... [22:28:54] oh I am now wondering, for the preferred bit, is there a comments field that adjoins it? [22:29:00] no! [22:29:16] There are some issues with the “preferred” schema, we should iron that out. [22:29:38] There is an edit comment however. [22:29:54] it’s just hard to correlate with the bit change itself. [22:30:34] Hmm, but I guess that makes sense because if the preferred bit is edited twice, those are two distinct reasons it was done. [22:30:40] You'd have to find the edit that flips the bit and then use that edit's edit summary. [22:30:46] exactly [22:32:21] 10Scoring-platform-team, 10JADE: Bikeshed: Explore alternative names for "judgment" in JADE - https://phabricator.wikimedia.org/T200365 (10kaldari) [22:32:30] 10Scoring-platform-team (Current), 10DBA, 10JADE, 10Operations, 10TechCom-RFC: Introduce a new namespace for collaborative judgments about wiki entities - https://phabricator.wikimedia.org/T200297 (10kaldari) [22:33:10] 10Scoring-platform-team (Current), 10DBA, 10JADE, 10Operations, 10TechCom-RFC: Introduce a new namespace for collaborative judgments about wiki entities - https://phabricator.wikimedia.org/T200297 (10kaldari) [22:37:11] harej: good idea to split out the users [22:37:19] they're different use cases! ;] [22:37:41] That’s a helpful insight [22:45:07] Cool, I was feeling like the use cases weren’t quite concrete but this list makes it clear that they were, just needed writing down. [22:45:50] 10Scoring-platform-team (Current), 10DBA, 10JADE, 10Operations, 10TechCom-RFC: Introduce a new namespace for collaborative judgments about wiki entities - https://phabricator.wikimedia.org/T200297 (10stjn) A small comment, sorry if I am asking in the wrong place: in the documentation I didn’t see anywher... [22:50:14] 10Scoring-platform-team (Current), 10DBA, 10JADE, 10Operations, 10TechCom-RFC: Introduce a new namespace for collaborative judgments about wiki entities - https://phabricator.wikimedia.org/T200297 (10awight) @stjn Thanks for bringing it up, we do see this kind of abandonment as a possibility. So far, ou... [23:00:02] * awight looks at the clock in alarm [23:00:19] hehe [23:00:53] 10Scoring-platform-team, 10ORES, 10Release-Engineering-Team (Kanban): Create gerrit mirrors for all github-based ORES repos - https://phabricator.wikimedia.org/T192042 (10mmodell) It seems git-lfs does not have any support for mirroring. I think that we will have to roll our own solution for mirroring from g... [23:03:29] /o\